Week 5

The aim of this lab was to complete the project. Over the past week we had refined the code for depth detection in the camera. We used MATLAB to develop the code, and ended up with an algorithm that can detect the floor, walls and obstacles. Below is the visual representation of the depth feed.

Our first problem of the day was that the tracker ball did not arrive, despite this being the 2nd time we had to order it. The 3D printed bracket was also still lost, so we just had to work with what we had for the body of the project. On the positive side however, the new motors we used did not seem to have the same issues as the old ones, so we were able to build the majority of the body of the robot, and also get the motor control board to function properly. However, due to time constraints, and not having the required equipment to build the body, we were unable to complete a full pathfinding algorithm, and therefore we are forced to control the robot with a keyboard. Although we were unable to fully complete the project as we intended, we think we learnt a lot from attempting this task. On to the bench inspection!

Week 4

By the end of this lab we wanted to have the motor control board fully integrated with the pi, and also have the body of the robot built. However, we ran into an issue at the beginning of the lab that made this impossible. Our 3D print was lost, so we were only able to build the parts of the robot that were laser cut. We were able to progress with the motor control board however, and successfully installed the required software onto the pi. However when we tested it, we found that the motors didn't seem to work. We eventually realised that the motor control board required a second power supply to function. Once we sorted this, we were able to use the motor control board to control the motor gear box and cause it to move in different directions at different speeds. We checked the outputs of the motor control board using the gpi-o test command, and they appeared to be correct, but one of the wheels did not appear to be moving correctly, and would only move sometimes. This appeared to be because the voltage was split unevenly across the two motors, and it wasn't receiving a high enough voltage to function.






Over the next week we will complete the camera code, and by the end of the final lab we hope to have our finished product!